What changed in Player One

Platform and admin updates from 17 to 28 June 2026, newest first. See the latest updates.

Release

One Automations home, clearer image models and steadier providers

This release tidies the everyday running of Player One. Flows, jobs and scheduled tasks now live together in one clear Automations area, image generation gets its own settings section, and provider connections behave more predictably - including clearer countdowns when a provider is temporarily at its limit.

New
Image models, properly organised
A dedicated image models section in Settings separates image-capable models from chat models, with clearer provider status. A pinned image model now tells you honestly when it fails instead of quietly switching.
Improvement
Steadier provider connections
Connecting and removing AI providers is more consistent, model lists refresh reliably after changes, and temporary provider limits show clear per-account countdowns.
Improvement
Scheduled tasks you can inspect
The scheduled task page now shows a proper run history with friendly times and transcript links, so you can see exactly what happened on each run.
Improvement
Hosted sites that stay up
Sites and apps that Player One hosts for you now restart more reliably and keep their public web addresses stable across platform updates.
New
Report a problem from inside Player One
A clearer in-app way to report platform problems, which gathers the useful debug context for you instead of asking you to collect it by hand.
Improvement
Smoother first-run setup
Getting started is shorter and more dependable, with better password manager support and a cleaner hand-off into your first real chat.
Release

Clearer chat links, email attachments, watchers and webhooks

This update makes the everyday experience smoother and gives Player One more ways to connect to the outside world. Links now render and work properly in chat and previews, you can send email attachments straight from the assistant, and new watchers and webhooks let Player One react to files, schedules and incoming events. Jobs are easier to continue and tidy up, and the whole interface is steadier under load.

Improvement
Links work properly in chat
Links in messages, tables and markdown previews now display and open correctly, with Back and Forward navigation inside previews.
Improvement
Smoother chat scrolling
The chat follows the response as it streams, scrolls to the bottom when you send, and stays put when you deliberately scroll up to read back, with no more jumps on keypress.
New
Send email attachments
Player One can now send files as email attachments straight from the assistant, with subject, CC and BCC, and support for send-only sending accounts.
Improvement
Smarter email recipients
Email sending now resolves plain addresses and saved contact names as recipients, with clearer messages when an action is not supported.
New
File and schedule watchers
Have Player One watch a file or poll something on a schedule, detect changes, and notify you, then switch itself off sensibly when it is no longer needed.
New
Incoming webhooks
Player One can create incoming webhook endpoints with their own access controls, so other tools and services can trigger work directly, all visible in one place.
Improvement
Continue and tidy up jobs
You can continue a finished job in the same run instead of starting a new one, revise completed work, delete individual runs, and the jobs menu is cleaner with team avatars.
New
Deep research and documents
A dedicated deep-research mode is suggested for research work, and document creation and conversion are handled more cleanly behind the scenes.
Improvement
Smoother onboarding
Setup now leads with your language preference and removes unnecessary confirmation pauses, so getting started is quicker and clearer.
Improvement
Steadier under load
Player One keeps using your AI account through brief provider slow-downs instead of setting it aside, gives clearer wording when something is interrupted, and tools you install now survive restarts.
Release

Smarter model routing, faster everyday use and a guided help panel

This update makes Player One quicker and lighter on busy systems, gives you more control over which model handles each task, and turns the in-product assistant into a proper guided help panel. Connecting a provider now discovers its models automatically, session history is easier to search and recover, and voice mode is steadier than ever.

New
Smarter model routing
Player One can now route different tasks and surfaces to better-suited models, with fallback choices ready, instead of relying on one broad default.
Improvement
Cleaner model picker
The model picker and settings screens were tidied up, with richer model details, clearer capability labels and better routing controls.
Improvement
Easier provider management
Provider connections are managed more cleanly behind the scenes, custom provider URLs are discovered more smoothly, and connection details are kept separate from the main config so changes are safer.
New
Multi-account onboarding
Onboarding now handles setups with more than one account, so different users, providers or services can each have their own connection profile.
New
Guided help panel
The in-product assistant has grown into a proper guided help panel across the admin area. It understands more pages and settings and follows where you are in the interface, so its help feels more relevant.
Improvement
Faster, lighter everyday use
Recent chats, sessions, model data, skills, memory monitors and health checks now load in smaller pieces instead of all at once, reducing slow starts and making the interface feel more responsive on busy systems.
Improvement
More resilient chats
Chat handling holds up better when a run is interrupted, the gateway restarts, or a message is still being accepted, keeping activity indicators in place and avoiding stray system text in the composer.
New
Voice replies start sooner
Voice replies can now start speaking sentence by sentence instead of waiting until the whole turn is finished, so calls feel more natural.
Improvement
Richer Telegram formatting
Telegram messages now support richer formatting, alongside cleaner WhatsApp linking and disconnect behaviour.
Improvement
Cleaner branding and fuller admin
Avatars and top-bar wordmarks load through a cleaner branding path, and the admin area gained fuller page coverage across jobs, memories, messaging, contacts, devices, alerts, usage and workspaces.
New
Auto-discovered models
Connecting a model provider now discovers its available models automatically, so you no longer need to supply a model list by hand in the common case.
Improvement
Easier session history & recovery
Chats are easier to find across agents with better filtering, date search and pagination, and deleting a chat now updates the sidebar straight away so there are no confusing leftover entries.
Improvement
Clearer tool access
Agents now get clear guidance when a tool is unavailable by policy, with system entries and guard cards, instead of repeatedly retrying the same failed action.
Improvement
Steadier voice mode
Voice mode resumes more reliably after a call or restart, picks up short speech more quickly, filters background noise more firmly, and gained a listening swirl, mute overlay and cleaner transcript.
Improvement
Better scheduled task visibility
Scheduled tasks gained richer list and run filtering, clearer session targeting, and smaller usability improvements so routine automations are easier to manage.
New
Sidebar agent switcher & voice controls
The sidebar now includes an agent switcher and voice call controls, making it easier to move between agents and in and out of voice mode.
Improvement
Smarter auto-titles
Chat auto-titles now wait until there is enough conversation context, reducing odd or premature chat names, and stray system notices are kept out of the conversation.
Improvement
Tighter, steadier admin operations
Admin sign-in is locked down with public self-registration switched off, and instance operations are rate-limited so bursts of work queue cleanly instead of overloading the system.